System to suit your character

Many Japanese believe that Japanese word-processing (JWP) technology has eliminated the difficulties associated with the thousands of multifunctional Chinese characters ( kanji ) that crowd Japanese texts. Nanette Gottlieb explains why in this study of the social and cultural aspects of JWP. She too thinks that JWP has "put paid to any remaining hopes of romanisation by solving the hitherto stubborn input-output problem". That I doubt, but I nonetheless found her book worthwhile.

Gottlieb does an excellent job of providing historical background and surveying mainstream Japanese opinion on JWP proliferation, but loses critical distance when she reports what Japanese say they feel when reading and writing. Psycho-linguistic research shows that intuitions about kanji cannot be trusted - for an overview, see J. Kess and T. Miyamoto's The Japanese Mental Lexicon . Also, Gottlieb's assessment of the non-affective impact of JWP has gaps. She discusses the gender divide among JWP users, but not contextual factors such as the problems of Japanese schools.

Instead of distinguishing functions that all word-processing systems share (for example cut and paste) from those needed to accommodate the Japanese writing system, Gottlieb examines design features much as an average Japanese consumer contemplating a purchase might.

Twenty years ago, Japanese documents were literally hand-crafted; JWP has changed that, boosting production and reducing the time needed for revisions and formatting. But the personal computer did at least as much, and faster, where alphabets are commonly used. According to the latest Economic Planning Agency "white paper", fewer than 15 per cent of Japanese had internet access in 1998 compared with almost 25 per cent of Singaporeans, and the rate of increase in internet users between 1995 and 1998 was slightly lower in Japan than in Malaysia. Gottlieb observes: "In 1997, only around 20 per cent of Japanese had PCs, compared to over 40 per cent in the United States and about 35 per cent in Australia." The "stubborn problem" has been transformed, but has it been solved?

One thing the PC did in the West was to make typing, once a specialist skill, ubiquitous. Kanji -conversion typing is slow and stressful compared with touch-typing. Users must constantly check for incorrect kanji and make style choices. It takes effort not to accept whatever the computer offers first, especially when writing extemporaneously. Manufacturers vie to cram more kanji and archaic forms into their look-up dictionaries, increasing the need to choose, but the homophonic errors generated during kanji selection are more difficult to catch later than all other kinds of kanji errors.

Gottlieb frequently cites romanisation advocate Yamada Hisao, but implies that his concerns are out of date. She calls romanisation "a dead issue", but paints it in either/or terms; actually, even if they try not to read it, most users type romanised Japanese to initiate input. Gottlieb acknowledges that the lack of effective standards complicates the searching of Japanese databases but does not probe the role of JWP manufacturers in blocking standardisation and inflating consumers' expectations, or explain how genuine global standards would prevent false characters ( mojibake ) from being sporadically introduced into Japanese files passed through multiple internet nodes.

What Gottlieb does say is that "the use of kanji on the internet... cannot be realistically described as a case of language nationalism". Surely that depends on how many non-Japanese learn the written language well enough to use JWP and understand Japanese postings? Unless romanisation becomes an alternative to customary script, at least in cyberspace, can Japanese become an international language? "There are now 14.2 million (internet) users each of Spanish and Japanese," notes Gottlieb, but there is only one Japan, and for the very reasons that Japanese native speakers find JWP helpful, the written form of their language remains a daunting challenge for most foreigners. According to a 1998 survey of American universities, where Japanese has risen to fifth place among foreign languages, Italian had more students, yet Spanish was an order of magnitude more popular than Italian.

Readers coming fresh to JWP will glean much valuable information from Gottlieb's study. JWP technology is transforming Japanese work and home life, but not necessarily because JWP systems do what advertisers claim. Greater scepticism on that score would have added depth to this otherwise admirable book.

J. Marshall Unger is professor of East Asian languages and literatures, Ohio State University, United States.